Date:  Fri, 25 Feb 1994 09:52:39 -0800

From:  David Harnick-Shapiro <david[AT SYMBOL GOES HERE]BUCKAROO.ICS.UCI.EDU>

Subject: Re: Phonetic/Phonemic E-Mail Alphabet



On Thu, 24 Feb 1994 10:27:04 EST, Wayne Glowka <wglowka[AT SYMBOL GOES HERE]MAIL.GAC.PEACHNET.EDU> writes:



> Since I just succeeded in getting into this group about twenty-four hours

> ago, I am ignorant of practice and protocol, but I have noticed a problem

> with transcription of ash, schwa, etc. Has anyone established a handy set

> of qwerty phonetic/phonemic symbols for general use?



I've included some replies to this question from times past (back when

the sci.lang newsgroup could actually be used :-) Back then, Klatt's

system was leading the pack; the references to other systems may be

of greater interest today.



------- Forwarded Messages



From: zacharsk[AT SYMBOL GOES HERE]cs.umn.edu (Ronald Zacharski)

Newsgroups: sci.lang

Subject: Re: ASCII Version of IPA???

Message-ID: <1990Aug10.184647.22476[AT SYMBOL GOES HERE]cs.umn.edu>

Date: 10 Aug 90 18:46:47 GMT

Organization: University of Minnesota, Minneapolis - CSCI Dept.





Klatt, Dennis H. Review of text-to-speech conversion for English.

 J.Acoust.Soc.Am 82(3):737-793.



describes 2 transcription systems. One is case insensitive and requires

one or two characters per symbol. Klatt states this system is

"nearly identical to ARPAbet" The other system is a case sensitive one

character system.



Here's the chart from that article.



  system

2 ch  1 ch  example

- ----------------------

 IY   i    beet

 IH   I    bit

 EY   e    bait

 EH   E    bet

 AE   [AT SYMBOL GOES HERE]    bat

 AA   a    pot

 AO   c    bought

 AH   ^    but

 OW   o    boat

 UH   U    book

 UW   u    boot

 RR   r    Bert

 AY   A    bite

 OY   O    boy

 AW   W    bout

 YU   Y    Butte

 AX   x    about

 IX   |    nieces

 P   p    pet

 B   b    bet

 T   t    tet

 D   d    debt

 K   k    kit

 G   g    get

 CH   C    Chet

 JH   J    jet

 M   m    met

 N   n    net

 NX   G    sang

 F   f    fed

 V   v    vet

 TH   t    thin

 DH   D    this

 S   s    set

 Z   z    zero

 SH   S    shed

 ZH   Z    azure

 W   w    wet

 YX   y    yet

 R   r    red

 L   l    let

 HX   h    head

 EN   N    button

 EL   L    bottle

 _   _    silence "phoneme" "





Ron Zacharski

University of Minnesota





------- Message 2



From: zwicky[AT SYMBOL GOES HERE]pterodactyl.itstd.sri.com

Newsgroups: sci.lang

Subject: IPA and ASCII

Message-ID: <32551[AT SYMBOL GOES HERE]sparkyfs.istc.sri.com>

Date: 10 Aug 90 18:46:40 GMT

Organization: SRI International, Menlo Park, CA 94025





This seems to be a question I've adopted...



Actual IPA-ASCII mapping standards don't exist, partly because ASCII

doesn't have enough bits in it, unless you use digraphs for what

are single characters in IPA. However, there is an existing standard

for the ASCII transliteration of *phonemic* transcription of

*English*, using one character per phoneme. (In a discussion one day

about least job-relevant skills, I concluded that one of mine was the

ability to transcribe into Klattese at my normal typing speed for

English. This, along with the ability to tell you whether or not a

piece of Code 39 barcode met the military spec, and if not why not,

using a specially printed version of a jeweler's loupe, has atrophied

over the years for obvious reasons.)



a = the vowel in cot

c = the vowel in caught

[AT SYMBOL GOES HERE] = the vowel in cat

e = the vowel in Kate

E = the vowel in pet

i = the vowel in Pete

I = the vowel in kit

o = the vowel in coat

u = the vowel in coot

U = the vowel in put

^ = the vowel in cut

x = schwa



b = the initial consonant in bat

d = the initial consonant in date

C = the initial consonant in chat

D = the initial consonant in that

f = the initial consonant in fat

g = the initial consonant in gate

h = the initial consonant in hat

j = the initial consonant in jilt

k = the initial consonant in cat

l = the initial consonant in late

m = the initial consonant in mat

n = the initial consonant in Nate

N = the final consonant in king

p = the initial consonant in pat

r = the initial consonant in rat

s = the initial consonant in sat

S = the initial consonant in, umm, shape

t = the initial consonant in tat

T = the initial consonant in thick

v = the initial consonant in vat

w = the initial consonant in wet

y = the initial consonant in yak

z = the initial consonant in Zach

Z = the medial consonant in leisure



DUs, DIs Iz taipd in kl[AT SYMBOL GOES HERE]tiz, alDo nat [AT SYMBOL GOES HERE]t Enihwer nir Dx spid D[AT SYMBOL GOES HERE]t ai

yUzd tu bi aibxl tu aten, [AT SYMBOL GOES HERE]nd kwait pasIbli wiT Ercrz In Dx

tr[AT SYMBOL GOES HERE]nskripSxn (xb[AT SYMBOL GOES HERE]ut hwIC ai du nat wiS tu bi Infcrmd). [AT SYMBOL GOES HERE]nd yEs, ai du

rili tck laik DIs, haipxrkcrEkt Do It mei bi.



One of the chief joys of knowing Klattese is that with only a few

logical additions, one can pronounce most erratically capitalized

words. (I assume, for instance, that X is a voiceless velar

fricative). NeXT comes out very satisfactorily, although NeWS doesn't,

really.



    Elizabeth





------- Message 3



From: jackson[AT SYMBOL GOES HERE]jabberwock.shs.ohio-state.edu (Michel Jackson)

Newsgroups: sci.lang

Subject: Re: Does an ASCII version of the IPA exist?

Summary: no such thing exists

Message-ID: <312[AT SYMBOL GOES HERE]jabberwock.shs.ohio-state.edu>

Date: 25 Feb 91 20:27:26 GMT

Organization: The Ohio State University, Division of Speech and Hearing Science





You have several choices:

(i) IPA. 1989.

"The IPA 1989 Kiel Convention Workgroup 9 report: Computer coding of

IPA symbols and computer representations of individual languages",

Journal of the International Phonetic Association v 19, n 2, pp 81-82.



(ii) Recommended by the IPA:

LaserIPA macintosh font from

Linguist's Software,

Box 580

Edmonds WA 98020-0580



(iii) Pullum, G. L. & Ladusaw, W. A. 1987. _Phonetic Symbol Guide_.

Chiicago: University of Chicago Press.



(iv) (Englsih only)

the "ARPABET". see, e.g.,

Klatt, D. H. 1987. "Review of text-to-speech conversion for English",

J. Acoustical Society America v. 82 n. 3, p. 737-793. See esp. p.

767-769.



    ---michel (jackson[AT SYMBOL GOES HERE]speech.mit.edu, jackson[AT SYMBOL GOES HERE]shs.ohio-state.edu)





------- Message 4



From: ederveen[AT SYMBOL GOES HERE]pttrnl.nl

Newsgroups: sci.lang

Subject: Re: Does an ASCII version of the IPA exist?

Summary: COST-CPA: ASCII version of IPA

Keywords: IPA, CPA

Message-ID: <1991Feb25.140631.65002[AT SYMBOL GOES HERE]pttrnl.nl>

Date: 25 Feb 91 13:06:31 GMT





In article <4862[AT SYMBOL GOES HERE]mindlink.UUCP>, Joel_Murray[AT SYMBOL GOES HERE]mindlink.UUCP (Joel Murray) writes:



> If an ASCII version of the IPA does exist could someone please e-mail the file

> to me?



There are several proposals for an ASCII version of the IPA.

One of them is COST-CPA, as described in report EUR 12023 EN,

European Research Project COST 209,

Man-Machine Communication by Means of Speech Signals,

page 236, by P. Molbaek-Hansen



groeten / salutojn,

Derk Ederveen

- -------------

Kath. Universiteit, Nijmegen / PTT Research NT-TWS, Leidschendam  - -  NL

ederveen[AT SYMBOL GOES HERE]lett.kun.nl       D_Ederveen[AT SYMBOL GOES HERE]pttrnl.nl  tel. +31/0 70 3323202

kunlt1::ederveen         dnlts::ederveen    ederveen[AT SYMBOL GOES HERE]hlsdnl5.bitnet

** esperanto(Lingvo) :- neuxtrala(Lingvo), internacia(Lingvo), dua(Lingvo). **





------- End of Forwarded Messages



--------

      David Harnick-Shapiro   Internet: david[AT SYMBOL GOES HERE]ics.uci.edu

Information and Computer Science   UUCP: ...!{ucbvax,zardoz}!ucivax!david

University of California, Irvine   BITNet: DBSHAPIR[AT SYMBOL GOES HERE]UCI